In the age of AI, who controls your vision?
Why I open-sourced a Revit-to-Blender renderer while everyone else is buying subscriptions — and what the word 'access' is really hiding.
Someone asked a fair question after I put Blendit out: why do we need this in the age of AI visualisation? Good question. The answer has a few parts. Start with the part people don’t expect: Blendit sits perfectly happily alongside AI. It feeds it, even. Copy the output of the visualiser into any image-generation model and let it work its magic. You could wire an open-source image model straight in as another layer, for free — that’s genuinely feasible, not a someday promise. So this was never a stand against AI.
The real question, for me, sits underneath that one: who controls your vision, and at what cost? Open-source tools let you take any part of the process and bend it to what you want. I built Blendit because I kept watching a free tool sit half-used while the industry paid more and more for subscriptions — for things that, as far as I could tell, open source already solved.
Did you catch the word that matters in that sentence? Access. Open a piece of 3D software like Blender and it drops you into a limitless void. Millions of possibilities, thousands of knobs for every one of them. Most architects and designers quit right there — not for lack of skill, but for lack of time to set up yet another digital environment from scratch. The paid tools solve exactly this. Slick interfaces built around what a designer actually needs, quick renders, no fuss, smooth and fast. That is real value, and it is fair enough that people pay for it.
Here is the bit that nagged at me, though. If the whole back end can be rebuilt with open-source tools that already exist, then what you’re really paying for is the UX — the layer that tames the void and hands you a button. And a UX layer can be open source too. That is all Blendit is: the slick front door, built on top of Blender, and given away.
There’s one more reason, less philosophical. Resilience. What happens when the internet is down and you still need to bang out a few renders for a client before the end of the day? Most cloud and AI tools will struggle (yes, some run offline — I know). A procedural visualiser doesn’t blink. It produces the exact same render offline as it does online, every time, deterministically. Your ability to do your job shouldn’t hinge on someone else’s server staying up, or their pricing staying reasonable.
So that’s the why. AI can render. The paid tools are genuinely slick. I just think the person doing the work should own the tools that do it — and open source is how you get there.